Results

cesm2.ssp245

Numerical evaluation metrics

  • Each row correspond an evaluation metric
  • Dark blue cell indicates the best model in the row.
  • Light blue cell indicates the second best model in the row.
Metric naive lstm cnn xgboost
diff_of_means −1.10 1.45 −5.27 7.53
ratio_of_sd 0.83 0.81 0.92 0.86
amplitude_ratio_of_means 0.54 0.58 0.63 0.53
maximum_error 0.30 0.34 0.36 0.28
ks_mean_on_coarse_res_with_extremes 0.59 0.33 0.32 0.44
rainy_hours_ratio_of_means 0.82 0.83 0.89 0.95
qqplot_mae 0.03 0.03 0.03 0.02
acf_mae 0.15 0.11 0.09 0.11
extremogram_mae 0.11 0.06 0.06 0.08

Time series of the first few days

Graphical evaluation metrics

  • Observations: cumulative distribution (Detrended QQ), dependency structure (ACF), and extremes (Extremogram, using 0.97 cut off).

  • Daily/Monthly gruped: Maximum hour histogram (only in daily variables), Amplitude distribution.

  • Hourly/Daily grouped: Mean profile per hour within days (or weeks witihin season)

  • In all plots, grey color represent the true ERA5 signal.

cesm2.ssp370

Numerical evaluation metrics

  • Each row correspond an evaluation metric
  • Dark blue cell indicates the best model in the row.
  • Light blue cell indicates the second best model in the row.
Metric lstm xgboost naive cnn
diff_of_means −1.32 4.57 −4.78 −8.99
ratio_of_sd 0.85 0.88 0.89 0.98
amplitude_ratio_of_means 0.59 0.53 0.55 0.64
maximum_error 0.34 0.27 0.32 0.36
ks_mean_on_coarse_res_with_extremes 0.23 0.39 0.38 0.25
rainy_hours_ratio_of_means 0.80 0.90 0.79 0.86
qqplot_mae 0.03 0.02 0.03 0.03
acf_mae 0.12 0.12 0.15 0.08
extremogram_mae 0.07 0.09 0.12 0.06

Time series of the first few days

Graphical evaluation metrics

  • Observations: cumulative distribution (Detrended QQ), dependency structure (ACF), and extremes (Extremogram, using 0.97 cut off).

  • Daily/Monthly gruped: Maximum hour histogram (only in daily variables), Amplitude distribution.

  • Hourly/Daily grouped: Mean profile per hour within days (or weeks witihin season)

  • In all plots, grey color represent the true ERA5 signal.

cesm2.ssp585

Numerical evaluation metrics

  • Each row correspond an evaluation metric
  • Dark blue cell indicates the best model in the row.
  • Light blue cell indicates the second best model in the row.
Metric lstm naive xgboost cnn
diff_of_means −1.30 −4.32 5.57 −8.86
ratio_of_sd 0.87 0.90 0.89 1.01
amplitude_ratio_of_means 0.59 0.55 0.53 0.65
maximum_error 0.35 0.31 0.27 0.37
ks_mean_on_coarse_res_with_extremes 0.29 0.47 0.41 0.28
rainy_hours_ratio_of_means 0.82 0.81 0.94 0.89
qqplot_mae 0.03 0.03 0.02 0.03
acf_mae 0.11 0.15 0.12 0.09
extremogram_mae 0.08 0.12 0.09 0.08

Time series of the first few days

Graphical evaluation metrics

  • Observations: cumulative distribution (Detrended QQ), dependency structure (ACF), and extremes (Extremogram, using 0.97 cut off).

  • Daily/Monthly gruped: Maximum hour histogram (only in daily variables), Amplitude distribution.

  • Hourly/Daily grouped: Mean profile per hour within days (or weeks witihin season)

  • In all plots, grey color represent the true ERA5 signal.

ec_earth3.ssp434

Numerical evaluation metrics

  • Each row correspond an evaluation metric
  • Dark blue cell indicates the best model in the row.
  • Light blue cell indicates the second best model in the row.
Metric xgboost lstm naive cnn
diff_of_means −18.59 −22.31 −24.89 −29.17
ratio_of_sd 0.93 0.89 0.88 0.99
amplitude_ratio_of_means 0.62 0.70 0.60 0.76
maximum_error 0.28 0.35 0.30 0.35
ks_mean_on_coarse_res_with_extremes 0.34 0.21 0.44 0.24
rainy_hours_ratio_of_means 0.74 0.67 0.66 0.71
qqplot_mae 0.04 0.05 0.05 0.04
acf_mae 0.15 0.13 0.18 0.11
extremogram_mae 0.08 0.05 0.10 0.04

Time series of the first few days

Graphical evaluation metrics

  • Observations: cumulative distribution (Detrended QQ), dependency structure (ACF), and extremes (Extremogram, using 0.97 cut off).

  • Daily/Monthly gruped: Maximum hour histogram (only in daily variables), Amplitude distribution.

  • Hourly/Daily grouped: Mean profile per hour within days (or weeks witihin season)

  • In all plots, grey color represent the true ERA5 signal.

mri_esm2_0.ssp245

Numerical evaluation metrics

  • Each row correspond an evaluation metric
  • Dark blue cell indicates the best model in the row.
  • Light blue cell indicates the second best model in the row.
Metric naive cnn lstm xgboost
diff_of_means 19.99 20.21 20.64 30.08
ratio_of_sd 0.71 0.74 0.71 0.71
amplitude_ratio_of_means 0.45 0.49 0.49 0.42
maximum_error 0.28 0.39 0.33 0.25
ks_mean_on_coarse_res_with_extremes 0.57 0.26 0.26 0.46
rainy_hours_ratio_of_means 1.00 1.14 1.03 1.20
qqplot_mae 0.04 0.05 0.04 0.04
acf_mae 0.17 0.10 0.11 0.13
extremogram_mae 0.16 0.10 0.08 0.13

Time series of the first few days

Graphical evaluation metrics

  • Observations: cumulative distribution (Detrended QQ), dependency structure (ACF), and extremes (Extremogram, using 0.97 cut off).

  • Daily/Monthly gruped: Maximum hour histogram (only in daily variables), Amplitude distribution.

  • Hourly/Daily grouped: Mean profile per hour within days (or weeks witihin season)

  • In all plots, grey color represent the true ERA5 signal.

mri_esm2_0.ssp370

Numerical evaluation metrics

  • Each row correspond an evaluation metric
  • Dark blue cell indicates the best model in the row.
  • Light blue cell indicates the second best model in the row.
Metric naive cnn lstm xgboost
diff_of_means 6.06 7.44 7.76 17.31
ratio_of_sd 0.79 0.82 0.79 0.78
amplitude_ratio_of_means 0.51 0.57 0.56 0.49
maximum_error 0.28 0.39 0.32 0.24
ks_mean_on_coarse_res_with_extremes 0.57 0.31 0.27 0.46
rainy_hours_ratio_of_means 0.87 1.01 0.92 1.06
qqplot_mae 0.03 0.03 0.03 0.02
acf_mae 0.15 0.08 0.10 0.11
extremogram_mae 0.11 0.06 0.05 0.08

Time series of the first few days

Graphical evaluation metrics

  • Observations: cumulative distribution (Detrended QQ), dependency structure (ACF), and extremes (Extremogram, using 0.97 cut off).

  • Daily/Monthly gruped: Maximum hour histogram (only in daily variables), Amplitude distribution.

  • Hourly/Daily grouped: Mean profile per hour within days (or weeks witihin season)

  • In all plots, grey color represent the true ERA5 signal.

mri_esm2_0.ssp434

Numerical evaluation metrics

  • Each row correspond an evaluation metric
  • Dark blue cell indicates the best model in the row.
  • Light blue cell indicates the second best model in the row.
Metric cnn naive lstm xgboost
diff_of_means 18.03 18.19 18.71 28.62
ratio_of_sd 0.78 0.73 0.73 0.73
amplitude_ratio_of_means 0.50 0.47 0.50 0.43
maximum_error 0.41 0.28 0.33 0.25
ks_mean_on_coarse_res_with_extremes 0.26 0.47 0.25 0.41
rainy_hours_ratio_of_means 1.14 0.98 1.01 1.19
qqplot_mae 0.05 0.04 0.04 0.04
acf_mae 0.09 0.15 0.11 0.11
extremogram_mae 0.08 0.13 0.08 0.11

Time series of the first few days

Graphical evaluation metrics

  • Observations: cumulative distribution (Detrended QQ), dependency structure (ACF), and extremes (Extremogram, using 0.97 cut off).

  • Daily/Monthly gruped: Maximum hour histogram (only in daily variables), Amplitude distribution.

  • Hourly/Daily grouped: Mean profile per hour within days (or weeks witihin season)

  • In all plots, grey color represent the true ERA5 signal.